We study the molecular mechanisms involved in assembly and function of translation initiation complexes involved in protein synthesis, using budding yeast as a model system owing to the powerful combination of genetics and biochemistry available to dissect complex pathways in vivo. The translation initiation pathway produces an 80S ribosome bound to mRNA with methionyl initiator tRNA (tRNAi) base-paired to the AUG start codon. tRNAi is recruited to the 40S subunit in a ternary complex (TC) with GTP-bound eIF2, to produce the 43S preinitiation complex (PIC), in a manner stimulated by eIFs -1, -1A, and -3. The 43S PIC attaches to the 5' end of mRNA, in a manner facilitated by cap-binding complex eIF4F (comprised of eIF4E, eIF4G, and RNA helicase eIF4A) and PABP bound to the poly(A) tail, and scans the 5 untranslated region (UTR) for the AUG start codon. Scanning is promoted by eIFs 1 and 1A, which induce an open conformation of the 40S, and by eIF4F and other RNA helicases that remove secondary structure in the 5' UTR. AUG recognition by tRNAi evokes irreversible hydrolysis of the GTP bound to eIF2, dependent on GTPase activating protein (GAP) eIF5, releasing eIF2-GDP from the PIC and tRNAi into the ribosomal P site. After joining of the 60S subunit, producing the completed 80S initiation complex, the eIF2-GDP is recycled to eIF2-GTP by guanine nucleotide exchange factor (GEF) eIF2B to prepare for the next round of translation initiation. Functions of eIF1, eIF1A, and eIF2b in recognition of start codon context. We and others showed previously that yeast eIF1, eIF1A and eIF2 function to block initiation at non-AUG triplets, but it was unknown whether they also enable scanning PICs to bypass AUGs in suboptimal sequence context. As in other eukaryotes, the yeast gene encoding eIF1 (SUI1) contains an AUG in poor context, which could underlie translational autoregulation of its expression. Previously, we and others described eIF1 mutations that increase initiation at UUG codons (Sui- phenotype). We recently obtained mutations with the opposite phenotype of suppressing UUG initiation in Sui- mutants (Ssu- phenotype) by selecting for suppressors of the lethality of the SUI5 mutation in eIF5. Remarkably, we found that Sui- mutations in eIF1, eIF1A, and eIF2b all increase eIF1 expression at the translational level in a manner dependent on the poor sequence context of the SUI1 AUG codon; whereas Ssu- mutations in eIF1 and eIF1A decrease eIF1 expression with the native, but not optimal, context present. Therefore, discrimination against weak context depends on specific residues in eIFs 1, 1A and 2b that also impede selection of non-AUGs, suggesting that context nucleotides and AUG act coordinately to stabilize the closed conformation of the PIC. The eIF5-CTD/eIF2b-NTD interaction promotes eIF1 release and stabilizes the closed/Pin state to enable AUG recognition. We showed previously that the C-terminal domain (CTD) of eIF5 mediates interactions with eIFs 1, 3, 5 and the TC to stabilize the multifactor complex (MFC) and promote binding of MFC components to the 40S subunit during PIC assembly. We and others also obtained evidence that mutations in the N-terminal domain (NTD) of eIF3c and eIF1 that disrupt its interactions with one another, the eIF2b-NTD, and eIF5-CTD within the MFC, reduce initiation accuracy. It was unclear, however, whether the Sui- phenotypes of such mutations involve impaired interactions with the eIF5-CTD. We also showed previously that the presence of excess eIF5 promotes eIF1 release and stable TC binding to the closed state of the PIC (dubbed Pin) and enables aberrant recognition of UUG codons (Sui- phenotype), but it was unknown if eIF5-CTD's interactions with MFC partners underlie its activity in promoting eIF1 release. Our collaborators in Gerhard Wagners lab recently determined the structure of the human eIF5-CTD by NMR spectroscopy and used CSP analysis to demonstrate that eIF1 and eIF2b-NTD binding sites overlap on the same face of eIF5-CTD. Importantly, we collaborated with Katsura Asano's group to show that eIF5-CTD substitutions sufficient to disrupt eIF5-CTD binding to eIF2b-NTD without loss of eIF5-CTD/eIF1 interaction (the KK mutation) confer an Ssu- phenotype, implicating the eIF5-CTD/eIF2b-NTD interaction in stabilizing the closed PIC conformation and start codon recognition. Moreover, our finding that the eIF5 KK mutation suppresses the Sui- phenotypes of particular eIF1 mutations that evoke premature release of eIF1 from 40S subunits at UUG codons, but not those (including eIF1-G107K) that block eIF1 release specifically at AUG codons, suggested that the KK mutation impairs the ability of eIF5 to promote eIF1 dissociation from the 40S subunit. Our collaborators in Jon Lorsch's group provided biochemical support for this deduction by showing that the KK mutation impairs the ability of excess eIF5 to reverse the antagonistic effect of eIF1-G107K on TC binding in reconstituted PICs. These findings are significant in showing that the eIF5-CTD/eIF2b-NTD interaction within the MFC promotes eIF1 release and stabilizes the closed/Pin state to enable AUG recognition. Specific domains in eIF4G bias RNA unwinding activity of the eIF4F complex towards duplexes with 5' overhangs Binding of eIF4F (eIF4E, eIF4G, eIF4A) to the mRNA cap structure recruits the helicase eIF4A to the 5' UTR of mRNA where it can promote 43S PIC attachment and scanning by unwinding secondary structures. In collaboration with Jon Lorsch's group, we obtained evidence that the three RNA binding domains in the scaffold molecule eIF4G impart a preference in yeast eIF4A for unwinding duplexes with single-stranded 5' overhangs, which should help to direct eIF4A to the 5' UTR and provide 5' to 3' directionality to scanning. Distinct roles of translation elongation factors 3 and 1A in modulating the function of eIF2a kinase Gcn2 Recycling of eIF2-GDP to eIF2-GTP by eIF2B is impaired under stress conditions, such as amino acid starvation, by phosphorylation of eIF2a by kinase Gcn2. The resulting reduction in TC abundance down-regulates general protein synthesis but specifically induces translation of transcription factor Gcn4, an activator of amino acid biosynthesis. The kinase function of Gcn2 is activated on translating ribosomes by uncharged tRNA occupying the A decoding site in a manner requiring Gcn2 interaction with Gcn1 bound to the same ribosomes. Translation requires cyclical association of eukaryotic elongation factors (eEFs) with the ribosome, and part of the ribosome binding domain in Gcn1 has homology to eEF3, suggesting that these proteins use overlapping binding sites on the ribosome. Indeed, we found that overexpressing the ribosome binding segments of eEF3 impedes eIF2a phosphorylation by Gcn2 in cells, consistent with eEF3 interference with Gcn1 regulatory function on the ribosome. Hence, we propose that the Gcn1-Gcn2 complex functions only on ribosomes with A-site-bound uncharged tRNA, because eEF3 does not occupy these stalled elongation complexes. We also obtained evidence that Gcn2 interacts with eEF1A in vivo in a manner independent of their separate ribosome interactions, and that the Gcn2-eEF1A interaction is diminished in amino acid starved cells. These findings suggested that eEF1A could function as a Gcn2 inhibitor and, consistent with this possibility, we showed that purified eEF1A reduced the ability of Gcn2 to phosphorylate its substrate, eIF2&#945;, but did not diminish Gcn2 autophosphorylation in vitro. Thus, it appears that both eEF1A and eEF3 negatively regulate Gcn2 activation on translating ribosomes by distinct mechanisms, limiting Gcn2 activation to stalled ribosomes harboring uncharged tRNA in amino acid-starved cells.